1980: How the Map Broke Open

Ronald Reagan won 489 electoral votes on 11/04/1980 and Republicans gained twelve Senate seats on the same night, taking the chamber for the first time since 1955. The American Presidency Project records Reagan at 43,904,153 votes and 50.7 percent against Jimmy Carter's 35,483,883 and 41.0 percent, with John Anderson taking 6.6 percent as an independent. Four days earlier the best-known public polls in the country had the presidential race inside the margin of error.

The standard account calls 1980 a late break to Reagan. The evidence points at a measurement problem underneath it. Gallup put 11 days of interviewing into the last 54 days of the campaign and none into the final three, so a race that may have been moving for a fortnight could only appear as a single jump at the end. A break looks sudden when the instrument is intermittent.

What would prove this wrong is evidence that continuous measurement in October 1980 would still have shown a flat race. The two campaigns were closer to continuous than anyone publishing, and both of their final numbers were close to the result.

What the 1980 election produced

The 1980 election produced a Republican presidential win of 489 electoral votes to 49, a Republican Senate majority of 53 seats, and a Republican House gain of 35 seats that still left Democrats in control of that chamber. The Senate's own party division table records the move from 58 Democrats and 41 Republicans in the 96th Congress to 53 Republicans and 46 Democrats in the 97th, with Harry Byrd Jr. of Virginia sitting as an independent throughout. The Office of the Historian of the US House records the House going from 278 Democrats and 157 Republicans to 243 and 192.

Senate figures from the United States Senate's party division table, which states that its counts reflect the division immediately following the election. House figures from the Office of the Historian of the US House. Seats contested in 1980 from Wikipedia's compilation of the 1980 Senate elections; the figures reconcile exactly with the Senate's own party totals.

Measure	Before	After	Change
Senate Republicans	41	53	+12
Senate Democrats	58	46	-12
House Republicans	157	192	+35
House Democrats	278	243	-35
Senate seats contested, Democratic-held	24	12	-12
Senate seats contested, Republican-held	10	10	0

The Senate before and after the 1980 election

Seats by party in the 96th and 97th Congresses, as the Senate's own historical table records them.

Republicans lost none of the ten Senate seats they were defending. Twelve Democratic senators who sought reelection were beaten, and how they were beaten matters. Nine lost general elections to Republicans. Three had already been beaten inside their own party: Donald Stewart of Alabama and Richard Stone of Florida lost Democratic runoffs, and Mike Gravel of Alaska lost his primary on 08/26/1980. All three seats went Republican in November.

What the public polls showed in the final two weeks of 1980

Public polling in the last two weeks of the 1980 campaign showed a close race, and the result was a 10-point Reagan win, which is a failure of cadence as much as of method. Andrew Kohut, who ran Gallup's methodology, published an account of the year the following summer. Gallup's own trend had Carter ahead 45 to 42 in the survey taken the weekend before the 10/28/1980 debate, and Reagan ahead 46 to 43 among likely voters in the final survey interviewed 10/30 to 11/01/1980. Gallup's published estimate, after allocating the undecided, was Reagan 47, Carter 44, Anderson 8.

Gallup's 1980 readings and the days it did not measure

Each bar is one Gallup field window, drawn to scale on a calendar running from 09/12/1980 to election day. The white space is the campaign nobody measured.

The jump people remember, from Carter plus 3 to Reagan plus 3, spans an interval in which Gallup interviewed on two days out of six. The 10/28/1980 debate falls inside one of the unmeasured stretches.

Final 1980 polls against the result

Reagan's lead over Carter in each organization's last poll of the campaign, against the result on 11/04/1980.

Warren Mitofsky of CBS News assembled every final poll of the cycle into one table for the same conference. It shows the Washington Post's last poll with Carter ahead by 3, the CBS News and New York Times poll with Reagan ahead by 1, Gallup at Reagan plus 3, ABC and Harris at plus 5, and NBC and the Associated Press at plus 6. The result was Reagan plus 10. Kohut was blunt about the size of the miss: he wrote that the 3.8-point gap between Reagan's actual vote and Gallup's final measure of it was more than twice Gallup's average error over the previous six elections.

The two campaigns were not surprised. The same Mitofsky table carries both private polls. Richard Wirthlin's firm, polling for Reagan on 11/01 to 11/03/1980, had Reagan ahead by 11, and Patrick Caddell's firm, polling for Carter on 11/03/1980, had Reagan ahead by 10. Both campaigns' own research beat every public poll in the field, and both had it essentially right.

Kohut wrote that Gallup's interviewing closed before the collapse of the hostage negotiations that weekend, and that had the organization continued polling through it, he was reasonably confident it would have picked up the remainder of the trend line the final survey had begun to capture. The hostages had been seized on 11/04/1979, as the State Department's own history records, so election day fell on the first anniversary. Carter himself named the anniversary publicity as a factor the day after the vote.

What Reagan's 1980 coalition was built from

Reagan won in 1980 by taking 56 percent of white voters and 56 percent of independents while holding Carter to 48 percent in union households. The Roper Center's compilation of the CBS News and New York Times election-day survey of 15,201 voters gives Reagan 51 percent of all voters against Carter's 41, and the group figures underneath that total are where the campaign's design shows.

How groups voted in 1980

Each candidate's share of the vote within the group, from the election-day survey of 15,201 voters.

Three of those rows are the whole argument. Union households went 48 to 45 for Carter, which is a Democratic win and a catastrophic one, because a Democratic presidential campaign in 1980 needed that figure in the sixties. Voters aged 22 to 29 split 44 to 44, with Anderson taking the rest. And 27 percent of voters who called themselves Democrats voted for Reagan.

That last number is the one worth sitting with. Carter took 11 percent of self-identified Republicans. Reagan took 27 percent of self-identified Democrats. A crossover of that size is not assembled in October by advertising. It comes from finding a group whose party identification and current voting behavior have already separated, and giving them a reason to act on a decision they have partly made.

The conditions helped. The Bureau of Labor Statistics puts inflation in 1980 at an annual average of 13.5 percent, and unemployment stood at 7.5 percent in each of September, October and November 1980, against about 6 percent a year earlier. The bank prime loan rate averaged 13.79 percent in October 1980 and 16.06 percent in November on the Federal Reserve's series. Those numbers do not tell a campaign what to say. They tell it which voters are available to be asked.

Why twelve Senate seats moved at once in 1980

Twelve Senate seats moved in 1980 because Democrats were defending 24 of the 34 seats on the ballot and lost half of what they were defending. All 34 contested seats were Class 3 seats in regular elections, with no special elections that year, and the two-thirds Democratic share of that class is the structural fact underneath the result. Republicans could not have lost more than ten seats in 1980 under any national environment, because ten is all they had out.

Margins in the nine seats where Democratic senators were beaten

Winning margin in each 1980 general election where a sitting Democratic senator was defeated by a Republican.

Exposure sets the ceiling. Candidate quality and money decide how much of the ceiling gets used, and this is where the Republican committee work of 1979 and 1980 belongs. John Heinz of Pennsylvania chaired the National Republican Senatorial Committee from 1979 to 1981. An internal committee document from 1980, held in Heinz's papers at Carnegie Mellon, is unusually direct about what the committee thought mattered most. It states that the most important factor in likely Republican Senate gains that autumn was not the money at all, and that the committee's greatest success of the previous two years was in encouraging outstanding individuals to run, many of whom gave up safe congressional seats to make hard races against entrenched incumbents.

The same document describes a research department created in 1979 that put every Senate floor vote since 1974 into a computer, sorted into more than 900 categories, so a challenger could be handed any incumbent's record. That is opposition research built as infrastructure and amortized across a whole class of races.

The money ran the same way. The Federal Election Commission's final report on 1979-80 party activity shows the Republican Senate committee with net receipts of $23.3 million against the Democratic Senate committee's $1.7 million. The direct contributions to candidates were nearly even: $414,893 from the Republican committee against $481,500 from the Democratic one. The Republican advantage sat in coordinated expenditures, $5.0 million against $589,316, and in where the money went, with over half of its contributions going to challengers and open seats.

Independent spending added a separate layer aimed at the same targets. The Commission's study of 1979-80 independent expenditures put the total above $16 million, with Jesse Helms's Congressional Club first at $4,601,069 and the National Conservative Political Action Committee second at $3,307,962. The same study records that 78 percent of independent money spent on Senate races that cycle was spent against a candidate, and NCPAC's largest Senate target was Frank Church of Idaho, who lost by 0.96 points.

What 1980 says about measuring a race in its final week

The gap between the public polls and the campaigns' own polls in 1980 was not a gap in skill. Wirthlin and Caddell were in the field later, they were asking a narrower question of a narrower universe, and they were not publishing, so they did not need a number they could defend to an editor. Those three differences are available to any campaign that chooses to buy them, and they are worth more in the last ten days than in the preceding ten months.

The second lesson is about ceilings. A party defending 24 of 34 seats is exposed in a way no message can repair and no advertisement can create. The 1980 Senate class was set in 1974, when those seats were last contested in a very good Democratic year, and six years later the bill came due in a bad one. The arithmetic is public years ahead of the election it decides.

The third is about where a late break shows up first. Reagan's move came in the last four days and no public instrument measured it, because no public instrument was still in the field. A campaign that stops measuring at the point the news gets loudest has chosen to be surprised.

Questions about the 1980 election

How many electoral votes did Reagan win in 1980?

Ronald Reagan won 489 electoral votes in 1980 to Jimmy Carter's 49. The American Presidency Project records Reagan at 43,904,153 popular votes and 50.7 percent against Carter's 35,483,883 and 41.0 percent, with John Anderson taking 6.6 percent as an independent candidate.

How many Senate seats did Republicans gain in 1980?

Republicans gained 12 Senate seats in 1980, moving from 41 seats to 53 and taking control of the chamber for the first time since 1955. The Senate's own party division table records the change. Republicans lost none of the ten seats they were defending that year.

Were the 1980 polls wrong?

The final public polls of 1980 understated Reagan's margin by between four and thirteen points. Gallup's published estimate was Reagan 47, Carter 44, and the result was 50.7 to 41.0. Andrew Kohut of Gallup later wrote that the 3.8-point gap on Reagan's own share was more than twice Gallup's average error over the previous six elections. Both campaigns' private polls were close to the result.

Which Democratic senators lost in 1980?

Nine sitting Democratic senators lost general elections in 1980: Birch Bayh of Indiana, Frank Church of Idaho, John Culver of Iowa, John Durkin of New Hampshire, Warren Magnuson of Washington, George McGovern of South Dakota, Robert Morgan of North Carolina, Gaylord Nelson of Wisconsin and Herman Talmadge of Georgia. Three more were beaten in their own party's primaries or runoffs: Donald Stewart of Alabama, Richard Stone of Florida and Mike Gravel of Alaska. All twelve seats went Republican.